Brush well and stay clean.

As most of the sensors in the next revision of the NMPP will be optical there will be one one major problem: bio-fouling. Algae will grow on the sensors and all kinds of little animals tend to find a comfortable place in between.

To keep the sensors clean it's necessary to wipe them at regular intervals. This can be done manually but a mechanical wiper that is automatically activated at regular intervals is more convenient. A general solution for this is using the Zebra-tech Hydro-wiper. A motorized brush with it's own control logic that is available in different variations for all kinds of different sensors. An excellent device BUT unfortunately too big for our application, and not capable of cleaning 5 sensors at the same time.

A device that much better suits our needs is the Mechanical Wiper as used (and created by) Turner Designs. Although it's actually just a part of their C6 Multi-Sensor Platform they also sell it as a separate item (or more as a spare-part for the C6 probably). It's slim, small motorized wiper with two brushes that would nicely fit the NMPP.
There's only one problem. Turner Designs will sell us the wiper as a separate item, they will not tell us what is inside or how to control it.
But let's face it, what could be inside ? There must be a DC motor, probably something between 5 and 12 V DC, some way to read the position of the brush (after all it always has to make just one rotation and never block a sensor) and maybe some control logic.
We decided to take the risk and just order them. After all, if we found no way to get the internal control logic to work we could always just control the motor manually.

Much to our surprise it was simpler than we ever expected. Opening the wiper reveals just a small motor with built-in gear box and a three wire sensor attached to part where the rotating axis passes through. Some measurements quickly reveal that the motor is a 5V type (4.4 ohm coil resistance) and that the sensor is probably a hall effect sensor with an open collector output. And that's it. All the visible wires are just directly connected to the 6-pin connector on the outside.
So we just connected the power pins to a 5V DC/DC converter that is switched by the Campbell datalogger and the hall-effect sensor to one of it's analogue inputs. And it works fine. We power the 5V DC, wait some fixed time so the hall effect sensor is off and then wait until it switches back on. Stop the motor, and that's it. It's almost too simple, as illustrated by the connection diagram...